I'm about to start analysing some of my results. I don't think I will need to use something like SPSS as I can use graphs and charts to show what I need to show, and the material I am using for comparison used only graphs and charts. Do examiners expect statistical analysis? There seems little point in doing it just for the sake of it - and part of me thinks that the more reader-friendly the final document is, the more it might get read, which is one of my goals, not much point in it being stuck on a shelf. So anyone just using basic maths techniques???

This is not my area of study, but I think this is something your supervisor should be advising you about. They'll know the field and the level of analysis required. And am not sure that reader-friendly is a criteria that examiners care about! Reading the articles that my potential examiners have written, dense almost seems to be better! ;-)

I'm just using basic maths but then I'm a humanities student! Seriously though I agree with Sue2604 that this is something your supervisor would be the best person to advise you on. Especially because they're going to be involved with picking examiners, and should know what sort of things any examiners would be looking out for.

======= Date Modified 20 Jul 2009 06:44:21 =======
Hi Joyce,



I agree with the others. Also, what statistical techniques have people used in similar analyses? - this will also help you. If you are not trained in statistics and a program such as SPSS, you should seriously consider if you should do complex statistical analysis.



I had used SPSS before, and was quite good as statistics. However, I decided to use logistic regression in one of my case study analyses. This chapter took me the best part of a year to complete because of a complete lack of support from my supervisors. Neither of them wanted to learn the specifics of what it is or how to do it. When I discussed it I was told "you sound like you know what you're talking about" - basically I blinded them with stats and it worked, but I was more concerned about what an external examiner would think. I took a logistic regression course, and although this was a social science course and mine was a lot more complicated as it was natural science, the tutor on the course was really helpful and I met with them seperately to go through my results. However, this took a while because they were not in my department. I was lucky that they were interested in its application in the natural sciences.



So, my advice is that if other people use complex statistics and you want to, seriously consider what support is available. If there is none, as in my case, I would not progress it further. Unless everyone uses these techniques, your external examiners will not be expecting it, and might just glaze over when they read it anyway, as my supervisors do.

I have used SPSS before, but the results I want to compare one of my set of results with are all just simple maths, percentages and the like, so I think I ought to do the same, as I want to show a trend across time.

I'm in humanities and my results need to be presented statistically but as they are basic - ie how many of a certain thing over a certain time period etc, I am just using graphs, pie charts and tables. I had considered using something more complicated but then decided it would not add anything the information I wanted to present. If this is all you need to show your results then stick to something simple, I say.